Back in 2005, the FAR 31.201-6 was revised to specifically allow contractors to use statistical sampling techniques as a method of identifying and segregating unallowable costs. Prior to that, the use of statistical sampling was not specifically prohibited but its use was so discouraged by Government auditors. As a result, many contractors did not consider it to be in their best interests to apply statistical sampling techniques to identify (and exclude) unallowable costs.
A properly designed statistical sampling plan allows contractors to reduce effort in its cost assurance process yet maintain a high degree of confidence that unallowable costs do not end up on billings to the Government.
When statistical sampling is used, FAR requires the following:
- the sample must be unbiased
- the sample must represent the universe from which the sample is drawn
- large dollar and high risk transactions must still be separately reviewed, and
- the entire process must permit audit verification.
FAR also wisely recommends that the sampling process be subject to an advance agreement between the contractor and the administrative contracting officer. Without an advance agreement, the burden of proof is on the contractor to establish that its methods meet the aforementioned criteria.
If you feel like you are drowning in your cost assurance processes, consider the use of statistical sampling.
Post a Comment